JP2006119831A - Information processor and information processing program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve a problem that the revision of application finds poor redundancy due to complex management for backup processing when data on a flush memory is managed in an address format, not in a file format in the backup for application data to the flush memory. <P>SOLUTION: A database 2 is constructed for associating an address 1 on a RAM (Random Access Memory) for the application data with backup data 4 on the flush memory. One variable ID (Identification) is provided to one variable. In addition, an information processor uses a means reconstructing the variable ID from the backup data to cope with the revision of application; and a means registering the variable so that the variable ID may not depend on the application. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an information processing apparatus and a program for backing up variable data and status data of an application program.

  When performing backup processing of status data and variables of application programs, as a backup medium, a peripheral storage device such as a hard disk or a DVD, a random access memory (hereinafter referred to as RAM) loaded with a battery for data storage, A storage flash memory is used. From the viewpoint of cost and maintainability, a method using a flash memory as a medium is excellent, and features such as large capacity and easy data reading.

  In the method using the flash memory as a medium, generally, when data is written to a certain address on the flash memory, the overwriting process cannot be performed on the address. In order to write data to the same address again, the entire physically defined unit composed of a plurality of bits called a sector is erased, and data can be written to the address again after the sector is erased.

  In addition, there is a limit on the number of data guarantees in rewriting the bits constituting the sector on the flash memory. Generally, rewriting a bit from 10 4 to 10 6 will result in a bad bit. There is a property that cannot be rewritten.

  When performing backup processing using such a flash memory, data is written sequentially from the address to be written, and once the data is written to the end of the sector, the data is again sequentially from the beginning of the next sector. In many cases, a sequential access method for writing is used.

  As a data management format, there are a format in which data is backed up as a file (hereinafter referred to as a file format) and a format in which data is directly written to a physical address of a flash memory (hereinafter referred to as an address format). .

  In the case of the file format, it is necessary to prepare an area called a directory portion for storing file information on the flash memory, and the frequency of rewriting the directory portion is likely to increase, which causes a problem that defective bits are likely to occur. There is also a problem that the structure is complicated and the data writing time and reading time tend to be long. However, this format is flexible in terms of changing the type and amount of application variables to be backed up, and has the advantage that the type and amount of variables can be changed regardless of the data structure of the application software.

  On the other hand, the address format does not require a special area such as a directory part in the memory, and the memory can be used effectively. In addition, since it is not necessary to create a special area, the data structure is simple. For this reason, the data writing time and reading time are often short. However, in order to cope with changes in the type and amount of variables, there is a disadvantage that the system designer needs to completely grasp the address on the flash memory.

  When designing an application that performs backup processing using the address format, the address of the backup data on the flash memory depends on the structure of the application software, so the system designer re-executes the backup program each time the application program is designed. In this case, it is necessary to fully grasp the address on the flash memory and consider the data writing efficiency and the robustness of the backup.

  For this reason, there has been a problem for the system designer that the design work of backup processing that is not related to the contents of the application increases, and the burden on the system designer becomes heavy.

  In backup processing using flash memory in information processing devices, address-type backup processing cannot be shared and reused as backup infrastructure, and there is a problem that the workload is often heavy for system designers. It was.

  In order to solve this problem, according to the present invention, the system designer can easily introduce the backup process into the application while the data read time and the data write time are short, and the flash memory write efficiency is improved. An object of the present invention is to provide an information processing apparatus or program that has high robustness in backup processing.

  Hereinafter, backup processing according to the present invention will be described with reference to the data structure of FIG.

(1) In the backup processing of the present invention, means for storing the correspondence between the position of the variable data 10 to be backed up on the RAM 100 and the save position on the flash memory 101 as a variable directory 2 on the RAM 100 with a variable ID attached. And means for saving the variable data 10 and the variable ID to a new save position as information for saving the variable data 10 and the variable ID based on a preset save condition, and assigning one variable ID to one application variable 10. To do.

  In this manner, a database called variable directory 2 storing the application variable location on the RAM 100 (hereinafter referred to as the RAM address) and the save location on the flash memory 101 (hereinafter referred to as the flash address) is installed on the RAM 100. The variable data 10 and the variable ID are set as one set and backed up in the flash memory 101.

  On the variable directory 2, an area (hereinafter referred to as a field) for storing various information of backup data is arranged, the RAM address is stored in the RAM address field 22, the flash address is stored in the flash address field 23, The variable ID is stored in the variable ID field 21. In this way, one field set 20 for storing a plurality of information is arranged for one variable ID.

  Further, a storage area (hereinafter referred to as an element) 4 relating to backup data is arranged on the flash address 101, and the backup data 16 and information relating to the backup data are stored in the element 4 (hereinafter this information is stored). The area 30 is referred to as a header), and data is written to the flash memory 101 using the element 4 as a unit of writing.

  The backup program monitors the variable data 10, the backup data 16, and the like, and when a prescribed backup condition is satisfied, the variable data 10 on the RAM 100 is set as a new element 4 at a new address on the flash memory 101. Store the variable ID in the variable ID header 31. At that time, the flash address field 23 on the variable directory 2 is updated to the new flash address value to which it has been moved.

(2) In the backup process of the present invention, when an application variable is an array variable 18, one variable ID is assigned to the array variable 18, and one index 19 is assigned to one element 19 of the array variable 18. (Hereinafter referred to as an array index), and each array element 19 is stored as an element 4 in the flash memory 101.

  The variable directory 2 of the array variable 18 is provided with a table (hereinafter referred to as an array table) 9 for storing the array index of each array element 19 and the flash address in association with each other.

  Fields are arranged on the array table 9, the array index is stored in the array index field 14, and the flash address of the array element is stored in the array element flash address field 15.

  Further, in each element 4 on the flash memory 101, each array element 19 is stored as backup data 16, and the variable ID of the array variable 18 is stored in the variable ID header 31, and the array index of the array element 19 is arrayed. Back up as one element 4 in the index header 32.

  When the prescribed backup condition is satisfied, the array element data 19 on the RAM 100 is stored as a new element 4 at a new address on the flash memory 101, and the variable ID and array index are stored. Also, the array element flash address field 15 is updated to the new flash address to which it has been moved.

(3) The variable directory 2 in the present invention is reconstructed based on the backup data 16 backed up in the flash memory 101 when the application program is started. The flash memory 101 is sequentially read from the head sector of the flash memory 101, and at the same time, the variable directory 2 is constructed for each variable ID. At this time, if there is an element 4 with the same variable ID, the contents of the variable directory 2 and variable data 10 with that variable ID are overwritten. Since new elements are sequentially stored in the flash memory 101, the contents of the variable directory 2 and the variable data 10 are synchronized with the latest backup data 16 by this overwriting process. The variable directory 2 is reconstructed once for the entire flash memory 101, and then from the first sequential sector to the next writing address.

(4) In the present invention, the save areas of the flash memory 101 are grouped. The variable data 10 is divided into groups according to the types, and backup processing is executed in parallel according to the number of groups set.

(5) Further, sector erasure of the flash memory 101 according to the present invention is performed on the oldest sector in which writing has been performed when the address to be written moves to a new sector. An active flag for determining whether to delete the element 4 is prepared for the element 4, an active flag field 24 for storing the active flag is arranged on the variable directory 2, and a variable ID having an activated flag is provided. If there is the latest backup data on the erase sector, the element 4 is copied to a new sector. Sector erasure is performed after all the elements 4 that should not be erased are moved on the erase sector.

(6) In the present invention, when power supply voltage drops (hereinafter referred to as power fail) occurs in the system hardware, data to be preferentially backed up (hereinafter referred to as power fail data) is set. In this case, the power fail data is always backed up every time before the sector erase is executed.

  According to the present invention, although it is data management in an address format, it can be used in common for different applications and hardware, a system designer can easily introduce backup into an application program, and provides a backup method that can be easily reused it can.

(1) Further, according to the present invention, by using the variable directory, the designer of the application program can routinely manage the address management of the flash memory in the backup process to the management via the variable ID, and the work load Can be reduced.

(2) Further, according to the present invention, the array element is individually backed up, so that the amount of data rewrite can be suppressed as compared with the case where the entire array variable is backed up, and the life of the flash memory can be increased.

(3) Further, according to the variable directory restructuring means of the present invention, it is not necessary to determine the latest data, and no special discrimination flag is required, so that the reconstruction can be performed with a simple data structure. .

(4) Further, according to the sector grouping means of the present invention, the frequency of data backup can be adjusted in accordance with the importance of data, and efficient sector utilization and life adjustment can be realized.

(5) Further, according to the sector erasing means of the present invention, sector erasure is performed only when there is no writable area in sector 3. For this reason, sector erasure can be made dependent only on the amount of data written. A very small sector erase frequency can be realized.

(6) Further, according to the power fail data backup means of the present invention, power fail data can be reliably backed up even when a power fail occurs during sector erasure of the flash memory.

  FIG. 1 is a configuration diagram of an information processing apparatus in which a flash memory, a variable directory, and variable data on a RAM according to an embodiment of the present invention are arranged. FIG. 3 shows a variable directory corresponding to one variable ID, It is a general-view figure with an element.

  The information processing apparatus 202 includes an application module 200 that performs processing other than backup of an application program, and a backup module 201 that performs backup processing. The central processing circuit 5 includes a flash memory 101, a RAM 100, an input device 6, and an output device. 7 is connected. In the application module 200, application variable data is arranged on the RAM 100 as the variable data list 1.

  In the backup module 201, a group of fields 20 for storing a plurality of information for each variable ID is arranged on the variable directory 2. As fields, a RAM address field 22 for storing a RAM address, a flash address field 23 for storing a flash address, a variable ID field 21 for storing a variable ID, and whether to move to a new sector 3 when erasing a sector are determined. An active flag field 24 or the like is used. Further, on the element 4 of the flash memory 101, a header set 30 for storing a plurality of pieces of information together with the backup data 16 is arranged. As the header, a variable ID header 31 that stores a variable ID, an array index header 32 that stores an array index, a size header 34 that indicates a data size, and the like are used.

  In this embodiment, a plurality of fields and headers are used for various processes described below. A type field 42 that describes whether the variable data is an array variable, an array number field 25 that describes the number of arrays in the case of an array variable, a bit size of the variable in the case of a single variable, and an array variable Includes a size field 43 that describes the bit size of the element, a registration flag field 27 that is registered by the application module to indicate whether the backup data is restored when variable data is registered and restored, and is erased even when erasing a sector A permanent flag field 28 for registering backup data to be continuously saved and a condition field 26 for describing the type of backup condition are used.

  It is also possible to provide a critical flag field 44 on the variable directory 2 and determine whether the access method when reading the variable data 10 is an access by interrupt prohibition, thereby realizing an interrupt prohibition function. In addition, a new sector field 45 for storing the number of the sector 3 in which the backup data is stored, an old sector field 46 for storing the number of the sector 3 in which the previous backup data is stored, and the previous sector field 46 are stored. An old flash address field 47 for storing the address of the element 4 where the backup data is stored, an error flag field 48 for describing an error when the backup data includes an error, and the like are stored. It may be used for specifying the previous backup data, specifying variable data, detecting change of variable data, and the like.

  Further, as the types of backup conditions specified from the condition field 26, the timing specified by the application is stored in the command field 41 and executed, or the backup condition specified in the period field 40 is executed in advance. It can be executed at a timing such as when it is executed based on the count in the counter field 29. Further, as a condition for performing backup, the application module 200 monitors each data specified by the RAM address and the flash address, and when the data contents are changed, the condition that the variable data 10 is backed up is added. May be. The variable data 10 specified by the RAM address and the information stored in the variable directory 2 and the array table 9 are backed up as new elements 4 to the addresses on the flash memory 101 according to the various conditions as described above.

  As described above, when executing the application program, the variable data 10 on the variable data list 1 is read or changed by the application. In the variable directory 2 of the variable ID, the variable data 10 on the variable data list 1 is referred to based on the RAM address stored in the RAM address field 22 and the backup on the element 4 based on the flash address field 23. The address of data 16 is obtained. Then, the previous period variable data 10 is stored in a new address on the flash memory 101 in accordance with a preset backup condition. At that time, the backup module 201 executes it separately from the normal processing of the application module 200.

  In the case of an array variable, each array element is backed up as a single element 4 to a new flash address. When the data indicated by one variable ID stored in the variable directory 2 is the array variable 18, the flash address field 23 on the variable directory 2 stores the RAM address of the array table 9 separately installed on the RAM 100. The array number field 25 is provided on the variable directory 2. In the array table 9, an array index for distinguishing each array element 19 is associated with an address on the flash address of each array element 19. Thus, by performing address management of the array elements 19 on the flash memory 101 one by one, for example, when data change occurs in only one element 19 of the array variable 18, the entire array 18 is backed up as backup data. Instead of backup processing as 16, only the array element 19 can be backed up as backup data 16.

  The data 19 on the RAM 100 of each array element can be specified from the RAM address specified by the RAM address field 22 on the variable directory 2 and the data size of the array element 19.

  In this example, the array variable 18 is assigned a variable ID to perform backup processing, but the array element 19 may be assigned a variable ID to perform backup processing.

  Next, a method of copying data into the flash memory 101 is shown in FIG. For each element 4, one backup data 16 and information corresponding to the backup data are stored as a header 30. First, the variable data 10 is stored sequentially from the top address of the top sector 3 of the flash memory 101. As shown in FIG. 2A, when there is an empty area larger than the size of data to be written in the sector 3 in which writing is performed, data is written into the empty area. On the other hand, as shown in FIG. 2B, when there is no empty area for the data size, the next sector area is set as a sector area for new writing.

  When the writing area moves to a new sector, an empty sector area on the flash memory 101 is searched sequentially from the first sector of the flash memory 101. As a result, if the sector area from the sector 3 to which the flash memory 101 has been written to the specified number of sectors is not an empty area, the sector written most recently is erased. When erasing, all the elements 4 included in the sector 3 to be erased are erased at once. As shown in FIG. 2C, after copying the element 4 to be stored without erasing from the sector area ahead of the specified number to a new sector, the data in the erase sector is erased.

  Thus, a prescribed number of consecutively erased sector areas and continuously written sector areas always exist on the flash memory 101.

  When the last sector of the flash memory 101 is erased and a new sector erase is required, the first sector area of the flash memory is erased. Then, the writing area and the non-writing area of the flash memory 101 are always circularly circulated, and the oldest variable data is backed up in the first sector of the circular continuous writing area. The element 4 to be backed up is written at the last address of the last sector.

  At the time of sector erasing, the element 4 that satisfies a certain condition is moved to a new sector 3. This element 4 includes not only the element 4 that is used by the application and registered to perform backup, but also the element 4 of backup data that is registered so as to always remain on the flash memory 101 without being erased. The element 4 has a registration flag field 27 for storing information that the element is registered in the variable directory 2 and a permanent flag field 28 for storing information that the element is always left without being deleted. The attribute flag / checksum header 35 for storing the permanent flag is arranged in the element 4 on the flash memory 101, and the state of the permanent flag is also saved on the backup data of the flash memory 101. A variable whose registration flag is ON and a variable whose permanent flag is ON are moved to a new sector 3 without being erased when the sector is erased.

  Next, the timing for performing the backup process will be described. When the backup is periodically performed, the backup module 201 is automatically and repeatedly called from the application in some form. On the backup module 201 side, the condition field 26 is arranged in the variable directory 2, the time is measured with a configuration in which the periodic field 40, the counter field 29, etc. used for counting time and performing periodic processing are arranged, and the application Backup processing is performed when the period registered in advance is reached.

  When it is desired to perform a backup process every time the variable data 10 is changed, the variable data 10 indicated by the RAM address field 22 of the variable directory 2 and the backup data 16 indicated by the flash address field 23 are sequentially compared, Backup processing is performed when the values are different.

  In addition, in order to arbitrarily specify the timing at which backup processing is performed on the application side, a command field 41 is simultaneously placed in the variable directory 2 and an instruction from the application is stored. The command field 41 is monitored, and backup is performed when backup is instructed.

  FIG. 4 shows the processing flow of the backup example according to conditions. Backup processing is performed with a condition at different timings for a plurality of application data. When the processing of the arithmetic circuit is transferred to the backup processing, the following processing is performed for all the variable IDs registered in the variable directory 2. First, in step 1, it is determined whether a backup instruction from the application is set in the command field 41. If the variable is a backup instruction setting, backup processing is performed in step 9.

  Even when there is no backup instruction setting, if the condition field 26 is read in step 2 and the backup is set to be performed when the variable changes, the variable continues to maintain the default value in step 3. If so, backup processing is not performed. If not, it is detected in step 4 by comparing the value designated in the RAM address field 22 of the variable directory 2 with the value designated in the flash address field 23 in step 4. If there is no change, backup is not performed, but even if there is a change, in step 5, in addition to the condition that the changed state is maintained for several turns, after satisfying that condition Backup processing is performed in step 10.

  Even when the backup condition is not when the variable is changed, the condition field 26 is read in step 6, and if the backup process is set to be performed periodically, the periodic backup process is performed. In step 7, it is determined whether it is a regular time. If it is a regular time, the process of step 8 is performed. In step 8, it is determined whether the variable continues to maintain the default value. If it is not maintained, backup processing is performed.

  In step 3 and step 8, setting is made so that backup processing is not performed if the variable continues to maintain the default value. If the specified value has been maintained since the beginning of the variable declaration, it is not necessary to perform backup, and the amount of backup processing can be suppressed by determining this condition. For example, if all variables are set to 0 when declaring variables, backup of variable data that has no backup data in the past and that has a value of 0 can be omitted.

  Further, in step 5, when the condition that the changing state is maintained for several turns is added as a condition, the application designer arbitrarily sets time as a backup condition.

  Also, data for which no backup condition is described in the condition field 26 is not backed up.

  Next, a process for recovering the value of the application variable at the time of starting the application will be described. When the application is started, the variable directory 2 is reconstructed from the backup data 16 and the variable data list 1 is reorganized. After the variable directory 2 is constructed, the backup data 16 of the variable ID of the data to be reused specified by the application is read into the variable data list 1, and the variable ID to be backed up is specified and the registration flag field 27 of the variable directory 2 is specified. Registered in

  In general, on the flash memory 101, a continuous written sector area and a continuous non-written sector area are provided in a circulating manner. Further, backup data with the same variable ID is written in the written sector area a plurality of times. Data is sequentially arranged, and the first data in the continuously written area is the oldest written data, and is stored in order of time. Therefore, the newest data is located in the rearmost element among the elements having the same variable ID.

  When the variable directory 2 is reconstructed, first, all backup data 16 is sequentially written into the variable directory 2 from the head address of the first sector of the flash memory 101 to the last address of the last sector. At the same time, the position of the rearmost element among the elements arranged sequentially is searched, and rewriting from the head sector to the searched element is performed again. As a result, the variable ID of the variable directory 2 is always overwritten with the newest data.

  By reconstructing the variable directory 2, the data backed up to the flash address 101 and the variable directory 2 are matched. Thereafter, it may be further matched with the variable data list 1 of the application, or registration of backup processing conditions may be received.

  Next, FIG. 5 shows a procedure for executing backup when viewed from the application side. By the backup module activation process from the application in step 30, the backup process in step 20 is activated. When the backup process is activated, the variable directory 2 is reconstructed in step 21. At the same time, information on the backup condition of the variable is registered in step 31 by the application program. If the variable ID of the variable registered at this time already exists in the variable directory 2, the variable is registered and the registration flag is turned ON, and the value of the backup data is a RAM indicated by the variable address in the variable directory 2. It is written in the address and used as variable data 10 of the application. If the variable ID does not exist in the variable directory 2, it is registered as data having a new variable ID in the variable directory 2 (step 22).

  Thereafter, when the internal processing is passed, the backup processing in step 23 is performed. The delivery of the internal process may be in the form of a subroutine of an application program, or may be in the form of a background process in which the internal process is delivered by a scheduler or the like. Invoking the backup process from the application program is normally only the above-described startup and registration procedures, but the backup instruction in step 32 used when backup is performed under conditions other than the backup conditions of the registered settings, The re-setting for changing the backup setting in step 33 can be called at will.

  In this way, the backup from the application program is executed only by activation and registration.

  With such a configuration, it is possible to leave data and variable IDs that are not desired to be deleted even when there is a change in application. Since there is no dependency on the application other than the registration of the variable ID at the time of starting the application, if any change is made to the application, a new variable ID can be registered for the new variable, and the change is redundant. Sexuality can be increased.

  Next, an example will be described in which, even if a power failure occurs other than in the end process and a flash data write error occurs, past data that is alternative is automatically read and handled without reading error data.

  When the backup module 201 is activated and a power interruption occurs while some data is being written, the data corresponding to the described variable ID may be error data. . When restarting, it avoids erroneously reading such error data into the variable directory 2 and reconstructing it, and reads backup data backed up before the error data without reading the error data. Since the variable directory 2 was reconstructed by sequential overwrite processing, if the error data can be determined, the system is automatically loaded from the latest backup data that does not cause an error by not reading it into the variable directory 2. It can be recovered.

  For this purpose, a size header 34 for storing the data size simultaneously with the normal backup data 16 and an attribute flag / checksum header 35 for storing the attribute flag and checksum are prepared in the element 4. It becomes possible. When describing the backup data, the element 4 is described in the order of data size, backup data, attribute flag, and checksum. By describing in this order, the attribute flag and checksum are not correctly written in response to the occurrence of a write error due to a power failure before describing the attribute flag and checksum. For this reason, it is possible to detect that an error has occurred in the middle of writing in the element 4 in which the attribute flag and the checksum are not correctly written, and to prevent reading error data.

  In addition, when it is necessary to always synchronize a plurality of variables, even if a certain variable data 10 is backed up normally, the backup is performed due to a power failure or the like before backup processing of another variable to be synchronized at the same time. If it is interrupted, it must be detected and the past state must be set again.

  This synchronization processing can be handled by providing the flash memory 101 with a synchronization flag element and a termination synchronization flag element. For example, a unique variable ID is assigned to the synchronization flag element and the terminal synchronization flag element so that it can be determined that it is a synchronization flag element. In addition, variable IDs of group members to be synchronized as data are described. A terminal synchronization flag element is also described as an element paired with the synchronization flag element. Here, when performing backup processing on the members of the variable group to be synchronized, the members are described so as to be sandwiched between the synchronization flag element and the end synchronization flag element. By describing in this way, the power-off in the middle of the description of the synchronization group will be interrupted if the synchronization flag element, the member described as data, and the terminal synchronization flag element are not described according to the stored order. You can detect what happened. In addition, when describing these elements, a route flag field 49 and a chain field 50 are arranged on the variable directory 2. In the route flag, it is described whether it is an element to be stored at the head of the variable elements to be described in synchronization, and the chain field 50 indicates the variable data 10 to be stored next to the element. deep. In the synchronous backup process, when the variable data 10 whose route flag is ON is detected, another variable directory 2 indicated by the chain field 50 of the variable directory 2 is searched, and further described in the chain field 50 from there. The variable directory 2 is searched one after another. After the chain is described as the data of the synchronization flag element, the variable data 10 indicated by each variable directory 2 is backed up, and finally the terminal synchronization flag element is described. In this way, synchronous backup can be realized.

  In general, it takes time to erase data in the flash memory 101, and if the power is cut during that time, the entire sector being erased becomes an indefinite bit string. In order to prevent the error data from being read, the non-write sector may be set so that two or more sectors 3 are always empty. In the above-described process of detecting a new sector to be written, it is set so that two or more non-write sectors are always continuous, and if there are no non-write sectors, at least one sector that should be present after the non-write sector 3 is deleted. By processing in this way, even if the sector 3 is entirely indefinite bits due to power interruption, the sector is always erased and data is not read from it. In addition, even if there is a power interruption in the data migration phase in order to perform sector erase after migrating valid data existing in the sector 3 to be erased, an error is caused by the power interruption countermeasure in the data writing described above. Data reading can be prevented.

  In addition, the backup process is a regular process that is executed in the background, but you may want to execute the backup process at a special timing such as power failure where the input voltage to the information processing device drops for some reason. is there. In addition, there is a case where it is desired to store a variable with a high update frequency and an important variable separately.

  In order to distinguish and save the power fail and the variable, it is possible to multiplex the sector group. For example, variable IDs 0 to FFF are set as A sector groups, 1000 to FFFF are set as B sector groups, and the others are divided as C sector groups, and backup modules 201 corresponding to the respective sector groups are executed in parallel. To do.

  Register the variable to be backed up in the interrupt program that is activated at the time of power fail, call the backup module 201 in the interrupt program, store the variable to be backed up in another sector 3 at the time of power fail, and respond to power interruption To do. However, in the case of a power failure during sector erasing, the sector erasing process takes a very long time, and the power may be cut off before performing the backup process. At this time, it is generally not possible to save (write) by interrupt processing. For this reason, if power fail data is set, the power fail data can be backed up without fail by reliably backing up the power fail data before executing sector erase. It becomes like this. For example, this backup is performed when the size of the free area of the sector 3 in which writing has been performed becomes smaller than a certain size, and the sector 3 to be erased is erased after the backup is completed.

  As described above, application variables can be restored from normal backup data without reading error data, and a highly robust backup process can be obtained.

Configuration diagram of information processing apparatus according to the present invention Figure showing the sector erase procedure Overview of variable directory operations corresponding to variable IDs Flow chart showing operation according to backup conditions Flow chart showing the operation when starting backup processing

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Variable data list 2 Variable directory 3 Sector 4 Element 5 Central processing circuit 6 Input device 7 Output device 9 Array table 10 Variable data on RAM 14 Array index field 15 Array element flash address field 18 Array variable data 19 on RAM Data of array elements on RAM 20 Variable directory field set 21 Variable ID field 22 RAM address field 23 Flash address field 24 Active flag field 25 Number of arrays field 26 Condition field 27 Registration flag field 28 Permanent flag field 29 Counter field 30 Element Header group 31 variable ID header 32 array index header 34 size header 35 attribute flag / checksum 40 Cycle field 41 Command field 42 Type field 43 Size field 44 Critical flag field 45 New sector field 46 Old sector field 47 Old flash address field 48 Error flag field 49 Route flag field 50 Chain field 100 RAM
101 Flash Memory 200 Application Module 201 Backup Module 202 Information Processing Device

Claims (12)

In an information processing apparatus that saves application program variables in flash memory as backup data,
Means for storing a correspondence relationship between a position of the variable on the RAM and a save position on the flash memory as a variable directory stored on the RAM with a variable ID;
Means for saving the data of the variable and the variable ID based on a preset save condition to a new save position on the flash memory as information for saving;
An information processing apparatus comprising: The variable directory includes an array table as means for specifying each array element of the array variable when the variable ID specifies an array variable,
The information processing apparatus according to claim 1, wherein the information to be saved includes an index representing the array element.   2. The information processing apparatus according to claim 1, further comprising means for scanning the flash memory and reconstructing the variable directory when the application program is started.   The information processing apparatus according to claim 1, wherein sectors of the flash memory are grouped, and the saving unit determines a group to save according to the type of the variable.   The flash memory is erased in units of sectors, and when there is data referred to by the variable directory in the sector to be erased, means for moving the data to a new save position on the flash memory The information processing apparatus according to claim 1, further comprising:   The flash memory is erased on a sector-by-sector basis. When erasing a sector, data that has been set in advance so as to be preferentially saved when the power supply voltage drops is stored in a new save position on the flash memory. The information processing apparatus according to claim 1, further comprising means for saving before erasing the sector. In an information processing apparatus that saves application program variables in flash memory as backup data,
A function of storing a correspondence relationship between a location of the variable on the RAM and a save location on the flash memory on the RAM as a variable directory with a variable ID;
Based on a preset save condition, the variable data and the variable ID are saved in a new save position on the flash memory as information to be saved, and a save function that is executed asynchronously with the application program,
A program for realizing on a computer. The variable directory has a function of specifying each array element of the array variable in an array table when the variable ID specifies an array variable;
A function for saving the information to be saved by including an index representing the array element;
8. The program according to claim 7, which is realized by a computer.   8. The program according to claim 7, which causes a computer to realize a function of scanning the flash memory and reconstructing the variable directory when starting the application program.   The program according to claim 7, wherein sectors of the flash memory are grouped, and the saving function causes the computer to determine a group to be saved according to the type of the variable.   The flash memory is erased in units of sectors, and when there is data referred to by the variable directory in the sector to be erased, a function of moving the data to a new save position on the flash memory The program of Claim 7 for making a computer implement | achieve.   The flash memory is erased on a sector-by-sector basis, and when erasing a sector, pre-set data to be preferentially saved when the power supply voltage drops is stored in a new save position on the flash memory. The program according to claim 7 for causing a computer to realize a function of saving before erasing a sector.

Images